The hidden cost of failed agent runs: one task may burn twice the credits on Codex
ToneAromatic178 · reddit · 2026-09-15
A Reddit user who tried Tencent's WorkBuddy (cheaper than Claude/Codex, with DeepSeek V4.1 Flash currently free) raises an overlooked issue: the hidden cost of agentic workflows. A first failed run plus a tweak and retry still costs money — he estimates the same task on Codex would have burned twice the credits. Failed runs, unnecessary retries, edits to unrelated files, and thrown-away generated code all add up fast, alongside the time cost. He asks the community how they manage this.
More from coding & agent
- Matt Shumer: I've redesigned my entire workflow around t3dotcodes — mattshumer_ · 2026-09-15
- OpenAI Agents Used 10+ Undisclosed Websites for Unauthorized Comms in Tests — eyishazyer · 2026-09-15
- Same prompt showdown: Claude vs ChatGPT vs Figma Make — Tegadesigns · 2026-09-15
- Andon Labs Launches Pion, Agents That Run Fully Autonomous Companies — infoxiao · 2026-09-15
- Photo of fridge whiteboard list becomes a full Sam's Club order via Muse agent — armand_ruiz · 2026-09-15
- OpenAI's Codex app adds official Arch Linux support via pacman; ChatGPT desktop hits Linux — OpenAIDevs · 2026-09-15